RESEARCH ARTICLE

Swimming by microscopic organisms in ambient water flow

M. A. R. Koehl Æ Matthew A. Reidenbach

Received: 26 February 2007 / Revised: 12 July 2007 / Accepted: 20 July 2007 / Published online: 21 August 2007

� Springer-Verlag 2007

Abstract When microscopic organisms swim in their

natural habitats, they are simultaneously transported by

ambient currents, waves, and turbulence. Therefore, to

understand how swimming affects the movement of very

small creatures through the environment, we need to study

their behavior in realistic water flow conditions. The pur-

pose of the work described here was to develop a series of

integrated field and laboratory measurements at a variety of

scales that enable us to record high-resolution videos of the

behavior of microscopic organisms exposed to realistic

spatio-temporal patterns of (1) water velocities and (2)

distributions of chemical cues that affect their behavior.

We have been developing these approaches while studying

the swimming behavior in flowing water of the micro-

scopic larvae of various bottom-dwelling marine animals.

In shallow marine habitats, the oscillatory water motion

associated with waves can make dramatic differences to

water flow on the scales that affect trajectories of micro-

scopic larvae.

1 Introduction

Studies of swimming organisms are usually done in still

water or in a flume in which unidirectional flow is adjusted so

that an animal swimming steadily upstream maintains its

position in the working section of the tank. In contrast,

organisms swimming in nature can be buffeted by unsteady

ambient water motions. The smaller or more weakly swim-

ming the organism, the greater the effect of environmental

water motion on its trajectory. When microscopic organisms

swim in their natural habitats, they are simultaneously

transported by ambient currents, waves, and turbulence.

Therefore, to understand how swimming affects the move-

ment of very small creatures through the environment, we

need to study their behavior in realistic water flow condi-

tions. However, measuring the swimming behavior of an

individual microscopic organism (e.g., instantaneous

velocity of the organism, beating of cilia or appendages)

requires high-magnification imaging of that organism.

The purpose of the work described here was to develop a

series of integrated field and laboratory measurements at a

variety of scales that enable us to record high-resolution

videos of the behavior of microscopic organisms exposed

to realistic spatio-temporal patterns of water velocities and

distributions of water-borne chemical cues that affect their

behavior. We used microscopic larvae of marine animals to

study the effects of ambient water flow on trajectories of

very small swimming organisms.

1.1 Swimming by microscopic planktonic larvae

of benthic marine animals

Many bottom-dwelling marine animals disperse to new

sites by producing microscopic planktonic larvae that are

dispersed by ocean currents. Where those larvae settle back

down onto the substratum can affect not only the popula-

tion dynamics of those species, but also the structure of the

benthic communities into which they recruit (reviewed by

Roughgarden et al. 1991; Olafsson et al. 1994; Rothlisberg

and Church 1994; Palmer et al. 1996). When competent

larvae (larvae old enough to be capable of undergoing

M. A. R. Koehl (&) � M. A. Reidenbach

Department of Integrative Biology,

University of California, 3060 VLSB,

Berkeley, CA 94720-3140, USA

e-mail: cnidaria@socrates.berkeley.edu

123

Exp Fluids (2007) 43:755–768

DOI 10.1007/s00348-007-0371-6

metamorphosis into the bottom-dwelling form) move from

the water column to the substratum, they pass through the

benthic boundary layer. A number of studies conducted in

flumes with unidirectional currents have shown that tur-

bulent flow in the benthic boundary layer affects the

delivery of larvae to the substratum (reviewed in Nowell

and Jumars 1984; Butman 1987; Eckman et al. 1990;

Abelson and Denny 1997; Crimaldi et al. 2002; Koehl and

Hadfield 2004; Koehl 2007). Many shallow coastal sites

where larvae recruit into benthic communities are exposed

to waves, yet our knowledge of how ambient water motion

affects rates of larval settlement has been based on studies

in steady currents. Boundary layers are thinner in waves

and shear stresses along the bottom are higher than in

unidirectional flow at the same free-stream velocity

(Charters et al. 1973; Nowell and Jumars 1984). However,

net horizontal transport across a habitat in back-and-forth,

wave-dominated flow is slow, even though instantaneous

velocities in waves can be high (e.g., Koehl and Powell

1994).

Experiments done in still water have shown that dis-

solved chemical cues can induce the larvae of many types

of benthic marine animals to undergo metamorphosis into

the bottom-dwelling form (reviewed by Hadfield and Paul

2001). These chemical cues are released by organisms

living on the substratum, such as adults of the same species

as the larvae, their prey, or bacterial biofilms. A few

behavioral studies of swimming competent larvae have

shown that these chemical cues can also induce downward

motion in still water (Hadfield and Koehl 2004) and in a

unidirectional ambient water current (Tamburri et al.

1996). However, the low magnification of the video records

of larval trajectories in those studies did not reveal whether

such downward motion was due to active swimming or

passive sinking, nor did those experiments reveal the spa-

tial distributions in the water of the chemical cues inducing

the changes in larval trajectories.

We used larvae of the nudibranch Phestilla sibogae and

of the tube worm, Hydroides elegans, both of which swim

with cilia, to investigate the effects of ambient water flow

on the motion through the environment of swimming

microscopic organisms.

2 Materials and methods

To investigate how swimming by microscopic organisms

interacts with ambient water motion to determine the

movement of the organisms in their natural habitats, we

studied how the larvae of benthic marine animals move

from the water column to the substratum to settle into

suitable habitats. We have focused on the larvae of the sea

slug, Phestilla sibogae, which settle onto the substratum in

response to a water-borne species-specific metabolite of

their prey, Porites compressa, the abundant coral that

forms reefs in shallow habitats in Hawaii (Hadfield 1977;

Hadfield and Koehl 2004; Koehl and Hadfield 2004). To

study the interaction of swimming larvae with ambient

water flow, we measured water velocities and mass trans-

port on a variety of scales to determine realistic conditions

under which to measure the swimming responses of indi-

vidual larvae.

2.1 Field measurements of water flow

Our first step in determining how ambient water flow

affects the motions of swimming larvae was to measure

water velocities in the field in the habitats in which the

larvae make their way from the water column to the sub-

stratum. For the larvae of Phestilla sibogae, we measured

water velocities above coral reefs in Kaneohe Bay on the

island of Oahu, HI (N21�270, W157�470).As is typical of Hawaiian reefs, the dominant coral

species at our study sites was Porites compressa. We

measured water velocities above P. compressa throughout

the tidal cycle, when water depth above the reef ranged

between 0.10 and 0.80 m. On the days when we measured

water velocities, water temperatures were 23–26�C, mean

wind speeds were 6.6–6.8 m s�1, and maximum wind

speeds were 9.2–10.8 m s�1 (weather station in Kaneohe

Bay of the Hawaiian Institute of Marine Biology, Univer-

sity of Hawaii).

Water velocities were measured using a Sontek SP-

AV10M01 Acoustic Doppler Velocimeter (ADV) with a

measurement volume of *0.25 cm3 and a sampling rate of

25 Hz. A cable from the ADV was run to a boat anchored

nearby, where the data were recorded on a laptop com-

puter. Water velocities were recorded for periods of 3 min

at heights of 4 and 8 cm above the surface of the reef. The

distance of the sampling volume above the reef was mea-

sured both by the ADV and by ruler (see Finelli et al.

1999), and these distances agreed in all cases. The ADV

was held in position by a rigid scaffolding placed not to

interfere with the flow being recorded.

Our field measurements revealed that the shallow

coastal habitats in which the larvae of P. sibogae must

make their way from the water column to a coral reef are

characterized by turbulent, wave-driven flow. Some

examples of our field water velocity data are shown in

Figs. 1 and 2. How are microscopic larvae transported in

such flow? How does such water flow disperse the dis-

solved chemical cues released by the coral P. compressa

that induce the larvae to settle onto the substratum? We

addressed these questions in a series of laboratory flume

experiments.

756 Exp Fluids (2007) 43:755–768

123

2.2 Wave-current flume

Our field measurements of water velocities above coral

reefs were used to design the water flow in a wave-current

flume (12.5-m long by 1.2-m wide) that could produce both

a mean current and surface waves (Fig. 3). A steady re-

circulating water current was produced by a centrifugal

pump commanded by a digital frequency controller. Waves

could be generated simultaneously by a paddle-type

wavemaker that was driven by a servo motor and linear

actuator (Pidgeon 1999). The wavemaker paddle produced

waves at a single frequency (which could be adjusted) that

propagated in the direction of water motion. A sloping,

broad-crested weir at the downstream end of the tank

minimized reflection of wave energy, since the flow over

the weir was supercritical.

A section of coral ‘‘reef’’ was constructed in the wave-

current flume from skeletons of the branching, reef-form-

ing coral Porites compressa (provided by M. Hadfield,

State of Hawaii collecting permit #1999–2005, after they

had been used in other experiments). Left-over coral

skeletons, rather than living corals, were used both to

minimize the impact of this study on wild populations of P.

compressa, and to avoid the difficulty of maintaining

healthy corals in the laboratory. Baird and Atkinson (1997)

found little difference between the overall drag, Re*,

roughness length scale, and mass transfer coefficients of

living P. compressa and of their skeletons, thus our use of

coral skeletons to study water flow over reefs is justified.

The skeletons of coral heads (typically about 15 cm tall)

were packed together tightly, as they are in the field, such

that the constructed reef completely covered the floor of the

test section of the flume (1.8-m long and 1.2-m wide). The

coral skeletons used in the flume had branch widths and

inter-branch spacings of approximately 1 cm (Reidenbach

et al. 2006). The total water depth in the flume for all

experiments was 40 cm, with the depth of water above the

canopy being on average 25 cm.

2.3 Measurements of water velocities in the flume

using Laser Doppler Anemometry

A Dantec two-component Laser Doppler Anemometer

(LDA), operated in forward scattering mode, was used to

measure streamwise, u, and vertical, w, velocities. The

Fig. 1 Comparison of horizontal velocities measured using LDA at a

height of 2 cm above the top of the P. compressa ‘‘reef’’ in the

laboratory wave-current flume (top graph), and measured using ADV

at a height of 2 cm above the top of a living P. compressa reef in

Kanehoe Bay, HI (lower graph). In both cases the mean water depth

above the top of the reef was *25 cm. The flume is diagrammed in

Fig. 3

Fig. 2 Comparison of the power spectra for vertical velocity

fluctuations measured using LDA at a height of 2 cm above the top

of the P. compressa ‘‘reef’’ in the laboratory wave-current flume

(black line), and measured using ADV at a height of 2 cm above the

top of a living P. compressa reef in Kanehoe Bay, HI (black circles).

These are data from the same flow records plotted in Fig. 1. In both

cases the mean water depth above the top of the reef was *25 cm

Fig. 3 Diagram of the wave-current flume (details given in the text).

Planar-laser induced fluorescence was used to image the flux of

Rhodamine 6G dye from the surfaces of the coral under a combined

wave-current flow. Examples of velocity measurements and a

turbulence spectrum used in this flume are shown in Figs. 1 and 2,

respectively

Exp Fluids (2007) 43:755–768 757

123

LDA optics were coupled with a laser (Coherent Innova

90), operated at a wavelength of 514.5 nm. The measure-

ment volume was 0.1 mm in the vertical and streamwise

directions, and 1 mm in the cross-channel direction.

Velocity measurements in the horizontal and vertical

directions were made by detecting the Doppler frequency

shift of laser light scattered by small particles moving with

the fluid. The LDA system was positioned by a 3-axis

motorized rail-bearing traverse with positioning in the

vertical direction to a precision of 50 lm. Velocities were

acquired above the top of the coral canopy (z = 0 defined at

the tip of the uppermost branch of the coral) at heights

z = 0.2, 0.4, 0.7, 1.0, 2.0, 4.0, 6.0, 8.0, 10.0, 15.0, 20.0,

30.0, 40.0, 50.0, 75.0, and 100.0 mm. Velocity measure-

ments at each height above the canopy were collected at

50 Hz for 30 min. The free-surface displacements of the

air–water interface in the flume directly over the LDA

measurement volume was simultaneously measured using a

capacitance wave height gauge (Richard Brancker

Research Ltd., Model WG-30).

2.4 Comparison of water flow in the field

and in the flume

Although water flow over coral reefs in Kaneohe Bay was

more variable than in the laboratory wave-current flume,

the small-scale features of the flow that should affect the

movement of larvae and of dissolved cues above a reef

were reproduced quite well in the flume. The peak veloc-

ities and periods of waves in the flume (Fig. 1a) fell within

the ranges of those measured in the field (Fig. 1b),

although the waves were more regular in the flume. While

large-scale, low-frequency flow features in the field could

not be replicated in the flume, the turbulence structure of

field and flume flow were very similar (Fig. 2).

2.5 Effects of waves on fine-scale flow

Our LDA measurements of velocity profiles along coral

surfaces in the flume revealed that the superposition of

waves onto a unidirectional current can have dramatic

consequences to water motion at the fine spatial scales

encountered by larvae swimming in the water near a coral

reef. Examples of such LDA measurements are shown for

the case of a unidirectional current (Fig. 4a) and a wave-

dominated current (Fig. 4b). The root-mean-squared (rms)

velocity for a unidirectional current over the coral

increased logarithmically away from the canopy, as

expected for a turbulent boundary layer (i.e., Gross and

Nowell 1983). In contrast, the profile of the rms velocities

for the wave-dominated case revealed drastically different

flow structure. A very steep gradient in rms velocities, as

well as in peak velocities, occurred very near the surface of

the coral. Due to the oscillatory nature of the flow, which is

caused by strong pressure gradients that oscillate with the

period of the wave forcing (Sleath 1987), water parcels

must be quickly diverted around the coral structure, caus-

ing accelerations in the flow and ultimately inducing much

higher velocities immediately adjacent to the coral rough-

ness elements (Lowe et al. 2005). While flow diversion

around the coral also occurred in unidirectional flows, the

flow directly adjacent to the coral was much slower than in

waves because a thicker boundary layer developed due to

the lack of an oscillating pressure in the flow field.

The higher velocities near the coral due to wave action

had the effect of increasing both the fluid shear stress

imposed on the coral and the turbulent motions near the

coral surfaces. Peak Reynolds stresses measured near the

surfaces of the coral under the oscillatory flow were

Fig. 4 Root-mean-squared (rms) water velocity profiles, calculated

from LDA measurements made above a coral ‘‘reef’’ in a wave-

current flume for a unidirectional current with a free-stream velocity

of 9.7 cm/s (a), and for a wave-dominated flow with a background

mean current (b). In this example, the wave period was 3s, and the

peak freestream horizontal velocity in the downstream direction

attained during each wave was 11.3 cm/s, while the peak horizontal

velocity near the corals was 12.4 cm/s

758 Exp Fluids (2007) 43:755–768

123

<u0w0>max/Urms2 = 0.04 ± 0.005 while that for the unidi-

rectional flow were 0.008 ± 0.001, indicating a five-fold

increase in turbulence due to waves. While these flow

dynamics may be small in scale relative to the depth of the

water column, they can have a dramatic effect on a

microscopic larva attempting to settle onto the reef.

2.6 Planar laser-induced fluorescence (PLIF)

measurements of chemical cues leaching

from corals

To investigate how realistic wave-driven water flow affects

the dispersal of the dissolved chemical cues that affect

larval behavior, we measured the fine-scale instantaneous

distributions of chemicals released from a reef in the wave-

current flume. Rhodamine 6G fluorescent dye was used as

an analogue for dissolved substances, such as larval set-

tlement inducer, released by the corals. This is justified

because the Schmidt numbers (Sc = {kinematic viscosity

of the fluid}/{molecular diffusivity of the dissolved sub-

stance}) for chemicals dissolved in water are quite high.

The Sc of Rhodamine 6G in water is 1,250 (Barret 1989).

Although we do not yet know the Sc’s of settlement

inducer released by corals, even if it were an order of

magnitude different from that of the dye, the fine-scale

patterns of concentration distribution of the two would be

quite similar because molecular diffusivity is so low rela-

tive to water’s kinematic viscosity.

To mimic the release of dissolved inducer from the

corals, we painted coral skeletons with a mixture of equal

volumes of a solution of Rhodamine 6G dye (500 ppm in

fresh water) and gelatin powder (Difco Laboratories Bacto-

Gelatin). A layer 1-mm thick of this dye-gelatin solution

was painted onto the surfaces of the coral skeletons. A strip

along the midline of the coral ‘‘reef’’ 1.2-m long (parallel

to the flow direction) and 0.20-m wide was painted in this

manner and allowed to set in air for 30 min. These coated

corals were then placed into the water in the flume. As the

gelatin slowly dissolved, the dye was released into the

water to simulate dissolved substances leaching from living

corals. For each experiment, coated corals could be

exposed to water flow in the flume for approximately

10 min before uneven wear of the dye coating along the

surfaces of the corals occurred.

Planar laser-induced fluorescence (PLIF) was used to

determine the fine-scale spatial and temporal structure in

the water above the reef of the concentrations of dye

released from the corals. By illuminating a thin slice of the

water column with a sheet of laser light (1-mm thick) that

caused the rhodamine dye to fluoresce, we could measure

concentrations of this analogue for inducer on a spatial

scale relevant to the ambits of microscopic swimming

larvae (Fig. 5). Our PLIF system consisted of a laser,

imaging optics to expand and focus the laser light, a

scanning mirror to produce a sheet of light, and a digital

CCD camera to record the dye fluorescence in the flowing

water. Laser light was emitted by an Argon-Ion laser

(Coherent Innova 90) at an output of 1 watt. The laser

beam was first expanded using a 3· laser expander (Melles

Griot) to minimize transmission losses and then focused

using a 2 m focusing lens. A light sheet was created using a

moving-magnet optical scanning mirror (Cambridge

Technology model 6800HP). As the dye passed through the

laser light sheet, the fluoresced dye was imaged with a

CCD camera (Silicon Mountain Design with 1024 by 1024

pixels and 12 bit resolution) fitted with a Micro-Nikkor

55 mm flat-field lens to reduce curvature effects at the

image edges. A filter on the receiving optics, with a center

frequency of 555 nm and bandwidth of 30 nm, was used to

remove laser and ambient light wavelengths, leaving only

emitted light from the fluorescing dye. Pixel brightness was

proportional to dye concentration (calibration procedures

described in Crimaldi and Koseff 2001). Raw images were

processed to remove biases in the data, including varying

pixel dark response, varying pixel response to fluorescence

intensity, slow background changes in pH and temperature,

lens and optics aberrations, and laser attenuation due to the

background dye concentrations, as described in Crimaldi

and Koseff (2001).

Each image (of an area in the flume 21 · 21 cm), was

exposed by a single laser scan with a total integration time

Fig. 5 Planar laser-induced fluorescence (PLIF) image of chemical

flux from the surfaces of two P. compressa coral heads (which appear

black at the bottom of the image) in the ‘‘reef’’ on the floor of the

wave-current flume. The inset image is a ·5 magnification of the

portion of the scalar field indicated by the white box. The white dot in

the inset indicates the size of a larva relative to the dye filaments. The

color scale bar indicates dye concentration, which is normalized by

Cs, the concentration along the surface of the coral. Mean flow was

from right to left with a mean current of 7.8 cm/s and a superimposed

wave with a period of 5 s and orbital wave velocity amplitude of

±11 cm/s

Exp Fluids (2007) 43:755–768 759

123

of 30 ms. The advection of dye during this integration time

was much smaller than the typical pixel dimension of

200 lm, thus ensuring accurate mapping of the scalar

structure onto the pixel. Images were collected at a rate of

10 Hz. Typically, 100–500 sequential images were taken

during each experiment sequence.

By using PLIF to examine the instantaneous spatial

distribution of dissolved inducer in the water moving above

a reef on the fine spatial scale relevant to a microscopic

larva, we learned that filaments of inducer swirl around in

inducer-free water. Therefore, we reasoned that as larvae

swim or sink through the water, they move into and out of

inducer filaments (see insert in Fig. 5), rather than

encountering a continuous diffuse concentration gradient

as has been assumed in models of larval settlement in

response to aromas from the substratum (e.g., Eckman

et al. 1994). How do brief encounters with settlement

inducer affect the swimming of larvae of P. sibogae?

2.7 Rapid reactions of larvae to encounters

with filaments of inducer

The larvae of Phestilla sibogae swim by beating cilia along

the edges of a two-lobed swimming organ, the ‘‘velum’’

(Fig. 6a). To determine how larval swimming is affected

by brief encounters with filaments of dissolved inducer, we

had to record the action of the cilia and the velum of

individual larvae as the animals moved into and out of

inducer filaments, hence the animals had to be viewed

using a microscope. Because the field of view of a

microscope is small, we could not follow the velar actions

of freely swimming larvae. Instead, we used videomi-

crography to record the actions of the velar lobes of

individual larvae tethered in a small flume (mini-flume)

that moved water past each larva at the velocity of water

motion relative to an untethered swimming larva,

*0.2 cm s�1 (Hadfield and Koehl 2004).

The Plexiglas ‘‘mini-flume’’ had a working section that

was small enough (3-cm wide · 3-cm deep · 14.5-cm

long) to permit a larva to be viewed using a microscope so

that ciliary beating and the position of the velum could be

discerned. We videotaped individual larvae (60 frames

s�1) using a SPI Minicam mounted on the ocular of a Wild

stereomicroscope. A steady flow rate of filtered sea water

through the flume was maintained by a constant-head tank,

arrays of screens were used to create a flat velocity profile

across the middle of the working section where the larva

was positioned, and velocity was adjusted by raising or

lowering the constant-head tank with a lab jack. The mini-

flume was designed to be a flow-through system so that

background levels of dye and chemical cues would not

build up over the course of an experiment. Water velocity

past a larva in the mini-flume was measured by videotaping

the movement of small neutrally buoyant particles carried

in the moving water. The microscope was focused on the

mid-line of the flume (where the larva was positioned) and

only particles in sharp focus were digitized and used to

calculate water velocities (using SCION Image software)

to avoid errors due to parallax.

An individual larva was tethered by using Vaseline1 to

stick its hydrophobic shell to the tip of a fine stainless steel

insect pin (0.24-mm diameter) and was held in a fixed

position in the mini-flume in the field of view of the

microscope. The pin and larva were gently lifted from the

larval culture dish and positioned with a micromanipulator

in the mini-flume so that the larva could ‘‘swim’’ into the

flow (i.e., the water flow relative to the tethered larva was

the same as the water flow relative to a freely swimming

larva) (Fig. 6a).

Tethered larvae were exposed to filaments of test solu-

tions (filtered sea water, or various concentrations of

chemical inducer released by the coral, P. compressa)

labeled with 0.05 or 0.1% fluorescein that were carried past

them in the flowing water (Fig. 6b) (Hadfield and Koehl

2004). These filament encounters were designed to mimic

the exposure to inducer filaments that a freely swimming

larva would encounter in the turbulent flow above a coral

reef (Fig. 5). We estimated the time course of encounters

with odor filaments by a larva above a coral reef using a

computer simulation of larval motions (due to swimming,

ambient waves, and turbulence) through the changing

concentration fields recorded in PLIF videos (e.g., Fig. 5)

(Koehl et al. 2007). Swimming velocities measured for

untethered larvae in still water in aquaria (Hadfield and

Koehl 2004) were used in these simulations. Our calcula-

tions indicated that a larva moving through the water

passes into and out of filaments of inducer, and that larvae

close to the reef encounter filaments more often than do

larvae higher above the reef (Koehl et al. 2007). We sim-

ulated swimming through an inducer filament in the mini-

flume by gently releasing (using a syringe pump, Sage

Instruments model No. 351) a narrow filament of test

solution into the water through a syringe needle (bore

diameter of 0.5 mm) positioned by a micromanipulator

perpendicular to the flow upstream from the larva. The

needle was higher in the water column than the larva, so

the larva did not experience the wake of the syringe. The

filament of test solution was carried downstream across the

tethered larva by the water flowing in the mini-flume, as

though the larva were swimming through the filament.

Video records of these fluorescein-labeled filaments

showed that before they reached a larva, the momentum

induced by the filament-releasing system was damped out.

We videotaped the instantaneous responses of larvae,

including cessation and resumption of beating of velar

760 Exp Fluids (2007) 43:755–768

123

cilia, and partial or complete retraction or re-extension of

the velum or foot. Using frame-by-frame analyses of these

video records, we measured (to the nearest 0.017 s) the

time lag between the onset or cessation of these behaviors

and the time when the edge of a filament encountered or

left the chemoreceptive organ of a larva (Hadfield et al.

2000). The mini-flume permitted us to expose larvae to

different temporal patterns of inducer encounters and dif-

ferent concentrations of the chemical cue. We found that

swimming competent larvae of P. sibogae did not respond

to fluorescein dye alone, but that they stopped beating their

cilia and retracted their velum into the shell when they

encountered inducer above threshold concentration, and

they re-expanded the velum and resumed ciliary beating

upon exiting an inducer filament (Hadfield and Koehl

2004). If they had not been tethered, the larvae would have

sunk through the water when the velum was retracted.

Our high-magnification records of what larvae did dur-

ing brief encounters with inducer filaments revealed that

they kept the foot protruded out of the shell when the

velum was retracted. This observation explains why larvae

moving downwards in coral inducer in still water traveled

more slowly than both swimming larvae and than sinking

dead, fully retracted larvae (Hadfield and Koehl 2004): the

high drag on the foot of a larva sinking in response to

inducer slows down its rate of descent. The foot of a

competent P. sibogae larva is sticky, thus a larva with its

foot extended from the shell can adhere to a surface on

which it lands (Hadfield and Koehl 2004; Koehl and

Hadfield 2004). Past analyses of larval settlement onto

surfaces have often relied on sinking velocities measured

for dead or anesthetized larvae (e.g., Butman et al. 1988),

but our Phestilla results illustrate the importance of

knowing the postures and behaviors of larvae before

assuming that living larvae move like passive, dead ones.

How do the instantaneous behavioral responses of larvae

of P. sibogae to brief encounters with filaments of coral

inducer affect their motion relative to a coral reef in nature?

To address that question, we must put microscopic larvae

(whose responses to inducer were measured in the mini-

flume, and whose swimming and sinking velocities were

measured in still water in aquaria, Hadfield and Koehl 2004)

back into the wave-driven ambient flow over a coral reef.

2.8 Simultaneous measurements of water velocities

(particle image velocimetry, PIV) and of scalar

concentrations (PLIF)

Microscopic organisms such as larvae are carried along by

the water in which they are swimming, Because they are so

small, such organisms are transported across the habitat by

Fig. 6 Side views of a larva of

Phestilla sibogae. a, b are

frames of a video taken of a

larva tethered in a small flume

in which water was moved past

the larvae at the same speed and

opposite direction as the

swimming velocity of the larva

(video made using a SPI

Minicam mounted on the ocular

of a Wild dissecting

microscope). The tether was a

fine insect pin (diameter

0.24 mm). c, d are diagrams of

larvae in the same postures as

shown in a, b, respectively.

When the tethered larva was

‘‘swimming’’ in filtered

seawater, its velum was

extended and its velar cilia were

beating (video frame in a and

diagram in c). When a tethered

larva encountered a filament of

inducer from P. compressa(labeled with fluorescien dye), it

stopped beating its cilia and

retracted its velum into the

shell, but left its foot protruding

out of the shell (video frame in

b and diagram in d)

Exp Fluids (2007) 43:755–768 761

123

net current flow, and are carried around locally in turbulent

eddies. Those turbulent eddies also carry the filaments of

chemical cues to which the larvae react (Fig. 5). The only

way that a microscopic organism can move relative to the

water and chemical cues around it is to swim or sink. In

contrast, the organism moves relative to the substratum

both by swimming or sinking and by being carried by the

ambient water flow. Therefore, to determine the temporal

pattern of inducer encounters by a larva or the trajectory of

a microscopic animal relative to the substratum, we must

simultaneously measure both the fine-scale velocity vector

field of the flow and the fine-scale patterns of concentration

of dissolved inducer in the water.

Instantaneous velocity vector fields of turbulent flowing

water can be measured using particle image velocimetry

(PIV; technique described in detail in e.g., Cowen and

Monismith 1997). A method which combines measure-

ments of planar laser-induced fluorescence and particle

image velocimetry can be used to simultaneously measure

inducer concentration and velocity structure in two

dimensions. For our PLIF imaging, we used a laser with an

output wavelength of light at 488 nm to excite fluorescein

dye (mean excitation at 490 nm) leaching from coral

skeletons, as described above for rhodamine. A sheet of

laser light (produced with a scanning moving-magnet

mirror, as described above) illuminated a thin slice through

the water column in the flume, and the fluorescein dye in

that slice emitted light at a mean wavelength of 520 nm.

We recorded the dye motions in the flume using a digital

camera (Redlake motionscope with 480 · 420 pixel reso-

lution) fitted with an optical bandpass filter of 520 ± 10 nm

(Andover Corp.) so that only emitted light from the fluo-

rescent dye was imaged. This laser sheet was scanned to

illuminate the imaging field every 0.02 s, with a wait

period of 0.02 s between each scan. A second laser with an

output wavelength of light at 532 nm was simultaneously

used for our PIV imaging. This second laser illuminated

silver-coated glass spheres (11 lm in diameter, Potter

Industries) carried in the water moving in the flume, but did

not excite the fluorescein dye. The second laser was also

pulsed at 0.02 s intervals, but only during the time periods

when the PLIF laser was not being scanned. The pulsed

laser passed through a 30� cylindrical lens to create a light

sheet that was aligned along the same two-dimensional

plane as the PLIF laser. Particle motions illuminated by the

PIV laser were recorded using a second Redlake motion-

scope digital camera (Fig. 7a). Both cameras were aligned

so that they imaged the same field. Images of particle

trajectories were analyzed with PIV software (MatPIV

1.6.1) to calculate a velocity vector in every subwindow

(16 by 16 pixels), and these vector fields were superim-

posed on the simultaneously recorded scalar concentration

field (Fig. 7b). Such data permits us to assess the

instantaneous water flow and inducer concentration

encountered by a microscopic larva at any defined position

in the portion of the water column that we have imaged.

2.9 Putting data from different scales together

to determine larval trajectories through

the environment

A computer simulation of the transport of larvae relative to

the substratum can be conducted using data measured at the

Fig. 7 Example of simultaneous PIV and PLIF measurements over

P. compressa coral in wave-dominated flow in a flume. The flow was

oscillatory with a mean freestream velocity (from left to right in the

image) of U = 10 cm/s. Images shown here were collected when the

oscillatory flow was beginning to reverse the current from right to left.

a Frame of a PIV video. b PLIF image of dye concentrations recorded

0.02 s after the image shown in a. The velocity vectors of water

motion that occurred during this interval are shown in blue

762 Exp Fluids (2007) 43:755–768

123

different spatial scales described above. Larvae can be

simulated by behavioral algorithms (based on microscope

observations on the scale of micrometers of larvae in the

mini-flume) and can be assigned swimming and sinking

velocities (measured on the scale of millimeters to centi-

meters s for freely swimming larvae in aquaria). These

simulated larvae can be placed randomly in the PLIF/PIV

data (scale of 0.1 mm’s to cm’s) recorded in a wave-flume,

in which water flow and coral reef structure measured in

the field on the scale of centimeters to meters was mim-

icked. A larva sinks or swims, depending on the

concentration of inducer in the pixel in which it is located

in the PLIF video frame. The vector sum of the swimming

(or sinking) velocity of a microscopic larva and of the

water velocity in the PIV subwindow in which the larva is

located predicts where that animal will be in the next frame

of the PLIF video of the scalar (i.e., inducer) field. By

repeating such calculations for successive frames of the

PLIF/PIV videos, the trajectories of larvae swimming and

sinking in the turbulent, wave-driven, inducer-laden water

above a coral reef can be determined.

2.10 Larval swimming near surfaces in water currents

and waves

When swimming animals encounter surfaces, their behavior

can be altered. For example, in still water, the swimming

behavior of microscopic larvae of the marine tube worm,

Hydroides elegans, changes if they touch a surface covered

with biofilm (The mechanisms responsible for these

behavioral changes are not yet understood.). Surfaces in

marine habitats are first colonized by biofilms of bacteria

and other microorganisms; then larvae of larger animals

recruit onto the biofilmed substratum (e.g., reviewed by

Koehl 2007). Videos of larvae of H. elegans in dishes of still

water showed that they continued to swim when over clean

glass surfaces, but spent more time crawling on the sub-

stratum than swimming above biofilmed surfaces (J.

Zardus, M. Hadfield, T. Cooper, M. A. R. Koehl, unpub-

lished data). Contact with biofilms in still water also induces

the larvae of H. elegans to undergo metamorphosis (Unabia

and Hadfield 1999). If these larvae are being carried past the

substratum by ambient water flow, how do contacts with

biofilmed surfaces affect their trajectories?

Hydroides elegans are abundant members of the ‘‘fouling

communities’’ that grow on man-made structures, such as

ships and docks, in warm-water harbors worldwide. To study

the trajectories of the microscopic larvae of H. elegans in

realistic flow conditions, we first measured water flow across

surfaces in Pearl Harbor, HI, where H. elegans are abundant.

Our measurements of water velocity profiles (using a small

electromagnetic flow meter, Marsh-McBirney Model 523)

adjacent to dock surfaces, showed that the flow oscillated due

to wind chop superimposed on slow ambient currents

(Fig. 8a).

We videotaped the behavior of competent larvae of H.

elegans near different substrata in the ‘‘mini-flume’’

described above. An array of screens upstream from the

working section was used to produce a velocity profile

similar to that measured in Pearl Harbor, and a plunger

upstream of the screens was used to superimpose velocity

oscillations on the net downstream water motion in the

flume to simulate the wind chop measured in the field

(Fig. 8b). Water velocity profiles were measured as a

function of time by digitizing (Matplotlib 0.83) videos

(60 fps) of the paths of neutrally buoyant marker particles

carried in the water.

Competent larvae of H. elegans were placed in the

upstream reservoir of the flume and were carried by the

moving water through the working section of the flume. By

working in the mini-flume, we were able to make high-

magnification videos in which individual larvae could be

followed (Fig. 9). The water was not recirculated through the

flume so that we only recorded the first encounters of larvae

with a test substratum. We videotaped the larvae (60 fps) and

digitized their trajectories using Matplotlib 0.83.

3 Results and discussion

3.1 Putting data from different scales together

to determine larval trajectories through

the environment

Our computer simulation of the transport of larvae

Phestilla sibogae relative to the substratum was an

Fig. 8 Water velocities measured as a function of time 2 cm above

surfaces covered with the tube worm Hydroides elegans in Pearl

Harbor, HI (a), and in the lab in the working section of the ‘‘mini-

flume’’ in which larval behaviors can be recorded

Exp Fluids (2007) 43:755–768 763

123

individual-based model that coupled behavioral algorithms

[measured for microscopic larvae exposed to realistic pat-

terns of encounters with coral odor (Hadfield and Koehl

2004)] with fine-scale patterns of time-varying instanta-

neous flow velocities and odor concentrations [measured

above a coral reef in a flume (Reidenbach et al. 2007)

exposed to wavy, turbulent water movement like that

measured in the field (Koehl and Hadfield 2004)]. Our

calculations revealed that the simple behavior of sinking

during brief encounters with odor filaments can enhance

the rate of larval settlement onto a reef by about 20%

(Koehl et al. 2007). Thus, we learned that the behavioral

responses of slowly moving microscopic larvae to chemi-

cal cues can affect their trajectories in the environment,

even in turbulent, wave-driven ambient water flow.

Is it worth the effort to measure fine-scale time-varying

velocity and odor distributions under field flow conditions?

Past analyses of larval settlement in response to odors

assumed a constant, diffuse concentration gradient of dis-

solved chemical cues from the benthos (e.g, Crisp 1974;

Eckman et al. 1994). When we used the diffuse time-

averaged odor concentration gradient measured above our

reef in the flume to calculate larval transport rates, we

overestimated larval transport by about 15% compared

with rates we calculated using the time-varying instanta-

neous filamentous pattern (Koehl et al. 2007). Many studies

of larval settlement assumed that larvae are transported in

the benthic boundary layer like passive, negatively buoyant

particles (e.g., Hannan 1984; Butman 1987). When we ran

our model with the assumption that larvae sink continu-

ously like passive particles, we over-estimated the rate of

transport of larvae to the substratum compared with rates

we calculated using the behavior measured for larvae of

Phestilla sibogae that sink only while in coral odor above a

threshold concentration (Koehl et al. 2007). Thus, our

calculations indicate that ignoring time-varying, fine-scale

flow and odor distributions or instantaneous larval

responses can lead to overestimates of the rates of transport

of larvae to the substratum.

3.2 Larval swimming near surfaces in water currents

and waves

H. elegans larvae in still water swim along helical paths

(Fig. 10a). Such helical swimming has been shown to

enable organisms in still water to navigate relative to odors,

light, and gravity (e.g., reviewed in McHenry and Strother

2003). However, we found that when H. elegans were

swimming in the miniflume in water flow like that

encountered near surfaces in harbors, they were carried

distances of a cm or more (i.e., �100 body lengths) relative

to the substratum for each turn of the helix (Fig. 10b, c)

(M. A. R. Koehl, D. Sischo, T. Cooper, T. Hata, and M.

Hadfield, unpublished data). Furthermore, when the oscil-

lations due to wind chop were added to the slow currents

typical of harbors, larval trajectories were more varied and

showed more vertical motion (Fig. 10c) than in flow

without the wind chop (Fig. 10b). The importance of

helical swimming in navigation by larvae carried by tur-

bulent or wave-driven ambient water motion has not yet

been determined.

Fig. 9 Frame of a video of competent larvae of the tube worm

Hydroides elegans swimming in wave-driven flow near a surface

covered with a biofilm in a small wave-current flume. The larvae,

which look like white dots, are about 100-lm long. The magnification

was chosen to achieve the largest field of view in which the larvae

were still big enough to be clearly visible

Fig. 10 Digitized trajectories of competent larvae of H. elegans.

Each colored line shows the path of an individual larva. Larvae in

still water swam along helical paths (left panel), but were carried

downstream in a unidirectional water current (middle panel), or

were carried in a slow current plus small waves (typical of wind

chop in harbors) (right panel). The larvae had more variable

trajectories with more vertical motion when waves were superim-

posed on the current

764 Exp Fluids (2007) 43:755–768

123

By producing small-scale flow in the miniflume that was

similar to water motion measured in the field, we were

also able to record the behavior of individual larvae of

H. elegans as they encountered the substratum under

realistic flow conditions (Koehl, Sischo, Cooper, Hata, and

Hadfield, unpublished data). We learned that when com-

petent larvae of H. elegans contacted the flume floor, they

appeared to ‘‘hop’’ along the bottom, touching down two–

three times per centimeter that they moved downstream.

The touchdown durations of larvae that contacted biofilmed

surfaces were longer than for larvae that contacted clean

glass, the duration of a larva’s contact with the substratum

was shorter in flowing water than in still water, and the

variation in touchdown time was greater in waves than in

unidirectional flow. Our results indicate flowing water and

contact with biofilmed surfaces have dramatic conse-

quences for the trajectories of larvae of H. elegans.

3.3 Ambient water flow affects the motion

of swimming microscopic organisms

Studies of swimming are often done in still water or in

unidirectional currents in flumes. However, when micro-

scopic organisms swim in their natural habitats, they

simultaneously are being transported by ambient currents,

waves, and turbulence. Therefore, to understand how

swimming affects the movement of very small creatures

through the environment, we need to study their behavior

in realistic water flow conditions. However, quantifying the

swimming behavior (e.g., ciliary beating, body trajectory)

of an individual microscopic organism requires high-

magnification imaging of that organism over time, which is

very difficult in their natural habitats or in a large flume. In

this paper, we have described a series of integrated field

and laboratory measurements at a variety of scales that

have enabled us to record high-resolution videos of the

behavior of microscopic marine larvae exposed to realistic

spatio-temporal patterns of water velocities and of chemi-

cal cues that affect their behavior.

We found that the oscillatory water motion associated

with waves can make dramatic differences to water flow on

the scale of microscopic larvae (Reidenbach et al. 2007).

Although shallow marine habitats are often exposed to

waves or wind chop, most flume studies of the effects of

water flow on larval settlement onto the substratum have

been done in unidirectional flow (reviewed in Abelson and

Denny 1997; Koehl 2007). Fortunately, the tools are now

available to produce in the laboratory more realistic water

flow, based on field measurements on the spatio-temporal

sales relevant to larval swimming.

Not only do ambient currents affect the trajectories of

microscopic organisms by carrying them across the habitat,

but water movement also can stimulate them to change

their swimming behavior. For example, some larvae curtail

their swimming in rapid water currents (e.g., lobsters,

Rooney and Cobb 1991; polychaetes, Pawlik and Butman

1993). Similarly, snail larvae were induced to sink when

high turbulence dissipation rates were produced in a tank

by oscillating screens (Fuchs et al. 2004). Some types of

larvae orient their locomotion relative to the direction of

ambient water flow or shear (e.g., barnacles, Crisp 1955;

bivalves, Jonsson et al. 1991; bryozoans, Abelson 1997).

Larvae that have landed on a surface can stay or they can

reject the surface and resume swimming (reviewed by

Krug 2006); the larvae of some species of barnacles resume

swimming after landing more often in rapidly moving

water than they do in slow flow (Mullineaux and Butman

1991; Jonsson et al. 2004; Larsson and Jonsson 2006). As

these studies illustrate, laboratory swimming studies con-

ducted in still water may not reveal ecologically relevant

behaviors for those animals that alter their locomotion in

response to water movement in the environment.

3.4 Swimming by microscopic organisms can affect

their transport by ambient water flow

Whether marine larvae are simply transported like passive

particles by moving water or actively seek suitable habitats

has long been debated (e.g., reviewed by Butman 1987;

Woodin 1991; Jumars 1993; Koehl 2007). Evidence has

been accumulating that, although microscopic organisms

such as larvae are small and swim slowly relative to ocean

currents, their locomotory behavior can affect where they

are transported by ambient water flow.

As described above, we found that the behavior of sea

slug larvae that are being carried in realistic wave-driven

water flow near surfaces can affect their transport to the

surfaces (Koehl et al. 2007). Similarly, the swimming of

bivalve larvae in unidirectional flow in flumes affects their

transport to the bottom. (Tamburri et al. 1996; Finelli and

Wethey 2003). In addition, studies of ascidian larvae,

which are large enough to be seen and tracked by inves-

tigators in the field under natural flow conditions, have

shown that larval swimming responses to light and gravity

affect their transport across the habitat and their settlement

onto the substratum (reviewed in Young 1990; Worcester

1994; Koehl et al. 1997).

Not only can the behavior of larvae affect their transport

across the benthic boundary layer, but their locomotion

also can affect their horizontal transport over long dis-

tances. Large-scale water movements in the ocean carry

marine larvae between sites and from offshore waters to

the coast (e.g., reviewed by Roughgarden et al. 1991;

Rothlisberg et al. 1995; Shanks 1985). The vertical position

Exp Fluids (2007) 43:755–768 765

123

of larvae in the water column can determine where they are

carried, since water at different depths can move in dif-

ferent directions. Thus, the vertical swimming or sinking

behavior of larvae can affect their horizontal transport

despite how slowly they locomote relative to ambient water

currents (e.g., Cronin and Forward 1986; Shanks 1985;

Epifanio et al. 1989; Bingham and Young 1991; Rothlis-

berg et al. 1995; Tankersley and Forward 1994; Tankersley

et al. 1995; Queiroga and Blanton 2005). Many lab studies

have documented the behavioral responses of larvae to

environmental cues (e.g., light, gravity, pressure, salinity),

as well as the endogenous rhythms of such behaviors and

the ontogenetic changes in swimming, and have related

these responses to where larvae are carried by currents

(e.g., Forward and Cronin 1980; Forward et al. 1995;

Tankersley and Forward 1994; Tankersley et al. 1995; and

others reviewed by Chia et al. 1984; Young and Chia 1987;

Young 1995). However, some larvae swim too weakly for

such mechanisms to be important to their dispersal (e.g.,

Stancyk and Feller 1986), while others are such strong

swimmers that they can actively move horizontally in spite

of ambient currents (e.g., crabs, Luckenbach and Orth

1992; lobsters, Katz et al. 1994).

4 Conclusions

Studies of swimming are often done in still water or in

unidirectional currents in flumes. However, the motion of

microscopic organisms swimming in their natural habitats

cannot be understood without considering how ambient

water flow affects their trajectories and the transport of

chemical cues that induce behavioral responses. In shallow

marine habitats, the oscillatory water motion associated

with waves can make dramatic differences to water flow on

the scale of microscopic organisms such as larvae.

Acknowledgments This research was supported by National

Science Foundation grant # OCE-9907120 (MK), Office of Naval

Research grant # N00014-03-1-0079 (MK), The Virginia G. and

Robert E. Gill Chair (MK), a MacArthur Foundation Fellowship

(MK), a Stanford Graduate Fellowship (MR), and a Miller Postdoc-

toral Fellowship (MR). We thank M. Hadfield for the use of facilities

at the Kewalo Marine Laboratory, University of Hawaii, for collab-

orating with us on work involving living larvae, and for providing

coral skeletons (State of Hawaii collecting permit #1999–2005). We

thank J. Koseff for the use of facilities at the Environmental Fluid

Mechanics Laboratory, Stanford University, and for collaborating

with us on wave-flume experiments, and M. Stacey for the use of

flume facilities in the Department of Civil and Environmental Engi-

neering, University of California, Berkeley. D. Sischo (supported by a

National Institutes of Health MBRS RISE grant) and T. Hata took the

minflume videos from which Figs. 9 and 10 were made. We are

grateful to R. Chock, T. Cooper, A. Faucci, N. George, S. Jackson,

and M. O’Donnell for technical assistance, and to G. Rangan for

making the diagrams in Fig. 6c and d.

References

Abelson A (1997) Settlement in flow: upstream exploration of

substrata by weakly swimming larvae. Ecology 78:160–166

Abelson A, Denny MW (1997) Settlement of marine organisms in

flow. Annu Rev Ecol Syst 28:317–339

Baird ME, Atkinson MJ (1997) Measurement and prediction of mass

transfer to experimental coral reef communities. Limnol Ocea-

nogr 42:1685–1693

Barret TK (1989) Nonintrusive optical measurements of turbulence

and mixing in a stabily stratified fluid. University of California,

San Diego PhD Dissertation

Bingham BL, Young CM (1991) Larval behavior of the ascidian

Ecteinascidia turbinata Herdman; an in situ experimental study

of the effects of swimming on dispersal. J Exp Mar Biol Ecol

145:189–204

Butman CA (1987) Larval settlement of soft-sediment invertebrates:

The spatial scales of pattern explained by active habitat selection

and the emerging role of hydrodynamic processes. Oceanogr

Mar Biol Ann Rev 25:113–165

Butman CA, Grassle JP, Busky EJ (1988) Horizontal swimming and

gravitational sinking of Capitella sp. 1 (Annelida: Polychaeta)

larvae: implications for settlement. Ophelia 29:43–58

Charters AC, Neushul M, Coon D (1973) The effect of water motion

on algal spore adhesion. Limnol Oceanogr 18:884–896

Chia F, Buckland-Nicks SJ, Young CM (1984) Locomotion of marine

invertebrate larvae: a review. Can J Zool 62:1205–1222

Cowen EA, Monismith SG (1997) A hybrid digital particle tracking

velocimetry technique. Exp Fluids 22:199–211

Crimaldi JP, Koseff JR (2001) High-resolution measurements of the

spatial and temporal scalar structure of a turbulent plume. Exp

Fluids 31:90–102

Crimaldi JP, Thompson JK, Rosman JH, Lowe RJ, Koseff JR (2002)

Hydronamics of larval settlement: The influence of turbulent

stress events at potential recruitment sites. Limnol Oceanogr

47:1137–1151

Crisp DJ (1955) The behavior of barnacle cyprids in relation to water

movement over a surface. J Exp Biol 32:569–590

Crisp DJ (1974) Factors influencing the settlement of marine

invertebrate larvae. In: Grant PT, Mackie AM (eds) Chemore-

ception in marine organisms. Academic, London p 177

Cronin TW, Forward RB Jr (1986) Vertical migration cycles of crab

larvae and their role in larval dispersal. Bull Mar Sci 39:192–201

Eckman JE, Savidge WB, Gross TF (1990) Relationship between

duration of cyprid attachment and drag forces associated with

detachment of Balanus amphitrite. Mar Biol 107:111–118

Eckman JE, Werner FE, Gross TF (1994) Modeling some effects of

behavior on larval settlement in a turbulent boundary layer. Deep

Sea Res 41:185–208

Epifanio CE, Masse AK, Garvine RW (1989) Transport of blue crab

larvae by surface currents off Delaware Bay, USA. Mar Ecol

Prog Ser 54:35–41

Finelli CM, Hart DD, Fonseca DM (1999) Evaluating the spatial

resolution of an acoustic Doppler velocimeter and the conse-

quences for measuring near-bed flows. Limnol Oceanogr

44:1793–1801

Finelli CM, Wethey DS (2003) Behavior of oyster larvae (Crassos-trea virginica) larvae in flume boundary layer flows. Mar Biol

143:703–711

Forward RB Jr, Cronin TW (1980) Tidal rhythms of activity and

phototaxis of an estuarine crab larva. Biol Bull 158:295–303

Forward RB Jr, Tankersley RA, De Vries MC, Rittshof D (1995)

Sensory physiology and behavior of blue crab (Callinectessapidus) postlarvae during horizontal transport. Mar Freshw

Behav Physiol 26:233–248

766 Exp Fluids (2007) 43:755–768

123

Fuchs HL, Mullineaux LS, Solow AR (2004) Sinking behavior of

gastropod larvae (Ilyanassa obsoleta) in turbulence. Limnol

Oceanogr 49:1937–1948

Gross TF, Nowell ARM (1983) Mean flow and turbulence scaling in a

tidal boundary layer. Cont Shelf Res 2:109–126

Hadfield MG (1977) Chemical interactions in larval settling of a

marine gastropod. In: Faulkner DJ, Fenical WH (eds) Marine

natural products chemistry. Plenum, New York, pp 403–413

Hadfield MG, Koehl MAR (2004) Rapid behavioral responses of an

invertebrate larva to dissolved settlement cue. Biol Bull 207:28–43

Hadfield MG, Meleshkevitch EA, Boudko DY (2000) The apical

sensory organ of a gastropod veliger is a receptor for settlement

cues. Biol Bull 198:67–76

Hadfield MG, Paul VJ (2001) Natural chemical cues for settlement

and metamorphosis of marine-invertebrate larvae. In: McClin-

tock JB, Baker BJ (eds) Marine chemical ecology. CRC Press,

Boca Raton,pp 431–461

Hannan CA (1984) Planktonic larvae may act as passive particles in

turbulent near-bottom flows. Limnol Oceanogr 29:1108–1116

Jonsson PR, Andre C, Lindegarth M (1991) Swimming behavior of

marine bivalve larvae in a flume boundary-layer flow. Evidence

for near-bottom confinement. Mar Ecol Prog Ser 79:67–76

Jonsson PR, Berntsson KM, Larsson AI (2004) Linking larval supply

to recruitment: flow-mediated control of initial adhesion of

barnacle larvae. Ecology 85:2850–2859

Jumars PA (1993) Concepts in biological oceanography. Oxford

University Press, New York, p 348

Katz CH, Cobb JS, Spaulding M (1994) Larval behavior, hydrody-

namic transport, and potential offshore-to-inshore recruitment in

the American lobster Homarus americanus. Mar Ecol Prog Ser

103:265–273

Koehl MAR (2007) Minireview: Hydrodynamics of larval settlement

into fouling communities. Biofouling 23:1–12

Koehl MAR, Hadfield MG (2004) Soluble settlement cue in slowly

moving water within coral reefs induces larval adhesion to

surfaces. J Mar Syst 49:75–88

Koehl MAR, Powell TM (1994) Turbulent transport of larvae near

wave-swept rocky shores: does water motion overwhelm larval

sinking. In: Wilson H, Shinn G, Stricker S (eds) Reproduction

and development of marine invertebrates. Johns Hopkins

University Press, Baltimore, pp 261–274

Koehl MAR, Powell TM, Dobbins EL (1997) Effects of algal turf on

mass transport and flow microhabitat of ascidians in a coral reef

lagoon. Proc 8th Int Coral Reef Symp 2:1087–1092

Koehl MAR, Strother JA, Reidenbach MA, Koseff JR, Hadfield MG

(2007) Individual-based model of larval transport to coral reefs

in turbulent, wave-driven flow: effects of behavioral responses to

dissolved settlement cues. Mar Ecol Prog Ser 335:1–18

Krug PJ (2006) Defense of benthic invertebrates against surface

colonization by larvae: A chemical arms race. In: Fusetani N,

Clare AS (eds) Marine molecular biotechnology, Springer,

Berlin, pp 1–53

Larsson AI, Jonsson PR (2006) Barnacle larvae actively select flow

environments supporting post-settlement growth and survival.

Ecology 87:1960–1966

Lowe RL, Koseff JR, Monismith SG (2005) Oscillatory flow through

submerged canopies: 1. Velocity structure. J Geophys Res 110,

Art. No. C10016

Luckenbach MW, Orth RJ (1992) Swimming velocity and behavior of

blue crab (Callinctes sapidus Rathbun) megalopae in still and

flowing water. Estuaries 15:186–192

McHenry MJ, Strother JA (2003) The kinematics of phototaxis in

larvae of the ascidian Aplidium constellatum. Mar Biol 142:173–

184

Mullineaux LS, Butman CA (1991) Initial contact, exploration, and

attachment of barnacle (Balanus amphitrite) cyprids settling in

flow. Mar Biol 110:93–103

Nowell ARM, Jumars PA (1984) Flow environments of aquatic

benthos. Ann Rev Ecol Syst 15:303–328

Olafsson EB, Peterson CH, Ambrose WG Jr (1994) Does recruitment

limitation structure populations and communities of macro-

invertebrates in marine soft sediments: the relative significance

of pre- and post-settlement processes. Oceanog Mar Biol Ann

Rev 32:65–109

Palmer MA, Allan JD, Butman CA (1996) Dispersal as a regional

process affecting the local dynamics of marine and stream

benthic invertebrates. Trends Evol Ecol 11:322–326

Pawlik JR., Butman CA (1993) Settlement of a marine tube worm as a

function of current velocity: interacting effects of hydrodynam-

ics and behavior. Limnol Oceanogr 38:1730–1740

Pidgeon EJ (1999) An experimental investigation of breaking wave

induced turbulence. Stanford University PhD Dissertation

Queiroga H, Blanton J (2005) Interactions between behaviour and

physical forcing in the control of horizontal transport of decapod

crustacean larvae. Adv Mar Bio 47:107–214

Reidenbach MA, Koseff JR, Monismith SG, Steinbuck JV, Genin A

(2006) The effects of waves and morphology on mass transfer

within branched reef corals. Limnol Oceanogr 51:1134–1141

Reidenbach MA, Koseff JR, Monismith SG (2007) Laboratory

experiments of fine-scale mixing and mass transport within a

coral canopy. Phys Fluids 19:075107

Rooney P, Cobb JS (1991) Effects of time of day, water temperature,

and water velocity on swimming by postlarvae of the American

lobster, Homarus americanus. Can J Fish Aquatic Sci 48:1944–

1950

Rothlisberg PC, Church JA (1994) Processes controlling the larval

dispersal and postlarval recruitment of Penaeid prawns. In:

Sammarco PW, Heron ML (eds) Coastal and estuarine studies.

American Geophysical Union, Washington, pp 235–252

Rothlisberg PC, Church JA, Fandry CB (1995) A mechanism for

near-shore concentration and estuarine recruitment of postlarval

Penaeus plebeius Hess (Decapoda, Penaeidae). Est Coast Shelf

Sci 40:115–138

Roughgarden J, Pennington JT, Stoner D, Alexander S, Miller K

(1991) Collisions of upwelling fronts with the intertidal zone:

The cause of recruitment pulses in barnacle populations of

central California [USA]. Acta Oecol 12:35–52

Shanks AL (1985) Behavioral basis of internal-wave-induced shore-

ward transport megalopae of the crab Pachygrapsus crassipes.

Mar Ecol Prog Ser 24:289–295

Sleath JF (1987) Turbulent oscillatory flow over rough beds. J Fluid

Mech 182:369–409

Stancyk SE, Feller RJ (1986) Transport of non-decapod invertebrate

larvae in estuaries: an overview. Bull Mar Sci 39:257–268

Tamburri MN, Finelli CM, Wethey DS, Zimmer-Faust RK (1996)

Chemical induction of larval settlement behavior in flow. Biol

Bull 191:367–373

Tankersley RA, Forward RB Jr (1994) Endogenous swimming

rhythms in estuarine crab megalopae: Implications for flood-

tide transport. Mar Biol 118:415–423

Tankersley RA, McKelvey LM, Forward RB (1995) Responses of

estuarine crab megalopae to pressure, salinity and light: Impli-

cations for flood-tide transport. Mar Biol 122:391–400

Unabia C, Hadfield MG (1999) The role of bacteria in larval

settlement and metamorphosis of the polychaete Hydroideselegans. Mar Biol 133:55–64

Woodin SA (1991) Recruitment of infauna: positive or negative cues?

Amer Zool 31:797–807

Exp Fluids (2007) 43:755–768 767

123

Worcester SE (1994) Adult rafting versus larval swimming: Dispersal

and recruitment of a botryllid ascidian on eelgrass. Mar Biol

121:309–317

Young CM (1990) Larval ecology of marine invertebrates: a

sesquicentennial history. Ophelia 32:1–48

Young CM (1995) Behavior and locomotion during the dispersal

phase of larval life. In: McEdward LR (ed) Ecology of

marine invertebrate larvae. CRC Press, Boca Raton, pp 249–

278

Young CM, Chia FS (1987) Abundance and distribution of pelagic

larvae as influenced by predation, behavior and hydrographic

factors. In: Giese AC, Pearse JS, Pearse VB (eds) Reproduction

of marine invertebrates. Blackwell, Palo Alto, pp 385–463

768 Exp Fluids (2007) 43:755–768

123